Chronic pain affects -100 million adults in the US, and is inadequately treated with drugs, that are often toxic and have side effects (e.g., addiction). Electrical stimulation in targeted nerve fibers is a promising new therapy, but has had suboptimal efficacy and limited long-term success as its mechanisms of action are unclear. Complementary therapies, such as acupuncture and massage that also involve neuromodulation as a mode of action, have not been quantitatively assessed. Critical to advancing pain therapy is a deeper mechanistic understanding of how a nociceptive signal is processed and modulated in spinal dorsal horn (DH), the first central relay station of nociceptive signaling. There are 3 major functionally distinct subsets of neurons in the DH that play different roles in pain transmission. Excitatory neurons and inhibitory neurons form important local pain circuitry that modulates activity of projection neurons that send ascending pain signals to the brain. It is critical to understand the specific roles for each neuron subset and the therapeutic actions of neurostimulation, tactile inputs, and drugs. For example, do they respond differently to different therapies? Can certain patterns of stimulation selectively inhibit or excite any subset neurons to maximize pain inhibition? These fundamental questions could not be easily addressed in a quantitative manner before this study. First, experimental barriers limit probing the DH to uncover the circuit topology, because it has been difficult to differentiate different subsets of DH neurons while simultaneously studying their physiological properties. Computational models of the DH, on the other hand, can predict how changes in sensory inputs influence pain transmission, but current models are hand-tuned, assume a fixed circuitry, nonlinear, high dimensional and thus intractable for sensitivity analysis - rendering a computational barrier. We will break these barriers and will construct a tractable data-driven computational model of the DH that enables powerful predictions on how different treatments alter neuronal activity in the DH. State-of-the-art electrophysiological techniques and powerful mouse genetic approaches will delineate the effects of sensory stimuli and stimulation on various subsets of DH neurons, and these data will be used to estimate the parameters and circuit topology of a mechanistic model of the DH. Model reduction will then be applied to generate a tractable characterization of the DH enabling sensitivity analysis. Developing and validating this innovative model will allow predictions that may differentiate various pain treatments and integrative approaches that can be readily tested in animals. RELEVANCE (See instructions): Chronic pain affects about 100 million adults in the US, but remains inadequately treated. Critical to advancing pain therapy is a deeper mechanistic understanding of how a nociceptive signal is processed and modulated in spinal dorsal horn (DH), the first central relay station of nociceptive signaling. We will combine state-of-the-art electrophysiological techniques and mouse genetic approaches with system identification tools to construct a tractable computational model of the DH that will enable powerful predictions on how different treatments alter neuronal activity in the DH.